George E. Griffin,6 and Thomas B. Nutman1. 1Laboratory of Parasitic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health ...
1199
Albendazole Treatment of Children with Ascariasis Enhances the Vibriocidal Antibody Response to the Live Attenuated Oral Cholera Vaccine CVD 103-HgR Philip J. Cooper,1 Martha E. Chico,4 Genevieve Losonsky,2 Carlos Sandoval,4 Ivan Espinel,4 Rajeshwari Sridhara,3 Marcelo Aguilar,5 Angel Guevara,4 Ronald H. Guderian,4 Myron M. Levine,2 George E. Griffin,6 and Thomas B. Nutman1
1
Laboratory of Parasitic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, 2 Center for Vaccine Development, University of Maryland, Baltimore, and 3Emmes Corporation, Potomac, Maryland; 4Department of Clinical Investigations, Hospital Vozandes, and 5Ministry of Public Health, Quito, Ecuador; 6St. George’s Hospital Medical School, London, United Kingdom
Because concurrent infections with geohelminth parasites might impair the immune response to oral vaccines, we studied the vibriocidal antibody response to the oral cholera vaccine CVD 103-HgR in children infected with Ascaris lumbricoides and investigated the effect of albendazole pretreatment on the postvaccination response. Children with ascariasis were randomized to receive either 2 sequential doses of 400 mg of albendazole or placebo. After the second dose, CVD 103-HgR was given, and serum vibriocidal antibody levels were measured before and 10 days after vaccination. Postvaccination rates of seroconversion were greater in the treatment group that received albendazole (P p .06 ). Significantly greater rates of seroconversion and geometric mean titer were observed in the albendazole group in subjects with non-O ABO blood groups. A significant association was observed between vibriocidal seroconversion rates and treatment group, suggesting that A. lumbricoides infections impair the immune response to oral cholera vaccine, particularly in subjects of non-O blood groups.
Helminth infections are estimated to infect 3 billion humans worldwide and are most prevalent in poorer regions of the developing world. Most of these infections are due to intestinal helminths (or geohelminths), infection with which has been associated with significant childhood morbidity, including stunted growth [1], poor cognitive performance [2], and malabsorption [1, 3]. Malabsorption resulting in macronutrient and micronutrient deficiencies is most closely associated with those geohelminths that inhabit the small intestines, namely Ascaris lumbricoides [1, 4], hookworm [5], and Strongyloides stercoralis [5], of which A. lumbricoides is the most prevalent [6]. We have postulated that the presence of geohelminths in the small bowel might not only affect absorption of nutrients but also may interfere with attachment of oral vaccines to M cells that overlie gut-associated lymphoid tissue at the mucosal surface. Received 28 April 2000; revised 30 June 2000; electronically published 8 September 2000. Informed written consent was obtained from a parent or guardian of each subject; the protocol was approved by the institutional review boards of the National Institute of Allergy and Infectious Diseases and by the Hospital Vozandes, Ecuador. Financial support: National Institutes of Health (NO1-AI-45251 to M.M.L.) and the Wellcome Trust (G.E.G.). Reprints or correspondence: Dr. Philip J. Cooper, LPD, NIAID, 4 Center Drive, Rm. 4/126, MSC 0425, NIH, Bethesda, MD 20892-0425 (pc102d@ hotmail.com). The Journal of Infectious Diseases 2000; 182:1199–206 q 2000 by the Infectious Diseases Society of America. All rights reserved. 0022-1899/2000/18204-0025$02.00
Poor immunogenicity of live oral vaccines is a frequent problem in studies conducted in developing countries that use doses that are highly immunogenic in study populations from North America or Europe. Poor immune responses after vaccination have been described for both viral and bacterial vaccines, including polio [7–9], rotavirus [10, 11], and the live attenuated oral cholera vaccine CVD 103-HgR [12, 13], which necessitates the use of increased dose and/or a larger number of doses administered. This phenomenon may partly explain occasional vaccine failures and disease outbreaks [14]. For example, a single dose of 5 3 10 8 cfu of CVD 103-HgR is able to induce seroconversion in 185% of healthy Europeans and North Americans [15–17] but in only 20% of Thai adults [12] and in 16% of Indonesian children [13]. Adequate seroconversion (e.g., 170%) required a log-fold increase in dose [13]. Within developing countries, the response rates to 5 3 10 8 cfu of CVD 103-HgR are closely associated with socioeconomic level, and vaccine failure is most apparent in areas where sanitation is inadequate and hygiene is poor [13, 18]. There are several possible reasons for the poor immunological performance of oral vaccines in poor populations, including the following: (1) background intestinal immunity, which may prevent the generation of a measurable systemic immune response [12]; (2) a state of relative immunodeficiency secondary to poor nutritional status; (3) interference with live vaccine attachment by intestinal flora [8, 12, 19]; and (4) the presence of geohelminth parasites that may interfere with the attachment of vaccine to the intestinal mucosa and/or the immune response
1200
Cooper et al.
generated. Previous studies have shown that the presence of A. lumbricoides is associated with damage to the intestinal villi [20, 21], and, although little is known of the cellular immune response to geohelminths, other helminth parasites are capable of suppressing the immune response to unrelated or bystander antigens [22, 23]. If geohelminth infections are capable of impairing the immune response to an oral vaccine, then it follows that curative anthelmintic chemotherapy might reverse this effect. Therefore, to investigate the potential impact of geohelminth infections on the immune response to an orally delivered vaccine, we performed a randomized, double-blind, placebo-controlled study in which subjects with A. lumbricoides infection were randomized to receive either albendazole or placebo before receiving a single dose of live oral cholera vaccine. We chose to use a live attenuated strain of classical 01 Vibrio cholerae as the experimental vaccine at a dose (5 3 10 8 cfu) that is safe and highly immunogenic in adults in industrialized countries [15–17, 24, 25] but poorly immunogenic in adults and school-aged children in developing countries [12, 13, 18].
Materials and Methods Study participants. Study-site selection was based on preliminary surveys of helminth prevalence and intensity and levels of Inaba vibriocidal antibodies in 3 different regions of Ecuador. Helminth infection prevalences and intensities were similar in all 5 sites, and the chosen study site had the lowest background levels of vibriocidal antibodies. Schools in the district of Pedernales in Manabi Province were visited, and the nature of the study was explained to the teachers and heads of families. Informed verbal consent was obtained to collect 2 consecutive stool samples, and a fingerprick blood sample was drawn to assess hematocrit levels and to determine blood group. Children who met all the following criteria were eligible to enter the study: 6–13 years old; infection intensity 11000 eggs per gram (epg) of A. lumbricoides on both of 2 consecutive stool samples; hemoglobin 110 g/dL; absence of severe concurrent illness, including clinical evidence of immunodeficiency; and negative urine pregnancy test if the subject was female and >10 years old. All children with evidence of intestinal helminth infection but who were excluded from the study were offered a single dose of 400 mg of albendazole. Study design. Albendazole is active against all intestinal helminths, and a single dose of 400 mg is estimated to cure 190% of A. lumbricoides infections [26]. We chose A. lumbricoides as the index infection because previous surveys in the same area have shown that it is the predominant intestinal helminth and because it resides in the small intestine and may alter the mucosa at this site [20, 21], which is also the primary site of oral vaccine attachment and immune stimulation. We decided to administer 2 doses of albendazole separated by 1 month to ensure effective cure, to allow recovery of the gastrointestinal mucosa, and to prevent the establishment of new infections due to migrating larvae. Because a lower dose (5 3 10 8 cfu) of live attenuated oral cholera vaccine CVD 103-HgR has been shown to be poorly immunogenic in underprivileged populations in developing countries [12, 13, 18],
JID 2000;182 (October)
whereas a higher dose (5 3 10 9 cfu) is able to overcome this deficit [12, 13, 18], we hypothesized that a relative impact of concurrent A. lumbricoides infection could be most clearly demonstrated at the lower vaccine dose. Each subject was assigned a number (1–233) and received either 2 200-mg tablets of albendazole (Zentel; SmithKline Beecham, London, UK) or 2 identical placebo tablets (SmithKline Beecham) from coded plastic containers (numbered 1–233). The same dose of albendazole or placebo was repeated after 30 days. Randomization lists were prepared by the Pharmaceutical Development Service, National Institutes of Health (Bethesda, MD), by using a randomized block design with a block size of 50. Doses for both time points were placed into the same coded container. On study day 37, all subjects received a single 5 3 10 8 cfu dose of live oral cholera vaccine CVD 103-HgR (Swiss Serum and Vaccine Institute, Berne, Switzerland) reconstituted in 100 mL of purified bottled water according to the manufacturer’s instructions. Successful vaccination was indicated by consumption of >95% of the vaccine suspension. Parents and children were instructed that the children should not eat or drink for 90 min before and after vaccination. All study recruits were offered a single 5 3 10 9 cfu dose of CVD 103-HgR and 400 mg of albendazole at the end of the study. Assessments. A case-report book was filled out for each participant by a trained interviewer (M.C.), and all assessments were performed in a double-blind manner. Each subject, parent, or guardian (as appropriate) was questioned for the following: relevant personal information (sex, age, address, etc.); past history of cholera vaccination; past history of severe gastrointestinal illness suggestive of cholera; recent and current gastrointestinal symptoms; and history of recurrent infections suggestive of immunodeficiency. Each subject was examined physically for clinical evidence of immunodeficiency and for measurement of height, weight, triceps skinfold thickness, and mid–upper arm circumference. All measurements were performed in duplicate. Each subject was questioned also for gastrointestinal symptoms on study days 30, 37, and 47. At the time of vaccination (study day 37), each subject was given a questionnaire written in Spanish to complete, which was used to describe potential gastrointestinal symptoms related to the vaccine. Sampling of subjects. Stool samples were examined by the modified Kato-Katz technique [27] (on study days 0, 30, 37, and 47) for assessment of infection with A. lumbricoides and Trichuris trichiura, and stool samples were preserved in 10% formol and subsequently were examined for the presence of hookworm and other intestinal helminth parasites by the formalin-gasoline concentration method [27] (on study day 0 only). Blood samples were taken on study days 0, 37, and 47. Blood was drawn into 5-mL SST Vacutainer tubes (Becton Dickinson, Franklin Lakes, NY) and centrifuged within 2 h of drawing, and the serum was extracted. Next, we placed aliquots of serum into cryotubes; they were stored and shipped in liquid nitrogen. Serum Inaba vibriocidal antibody levels were measured at the Center for Vaccine Development, University of Maryland (Baltimore), by use of the microtiter method and expressed as reciprocal titers [28]. Immunological studies were run in a blinded fashion, using coded samples. Statistical analysis. With a 2-tailed test and significance set at .05, the study was designed to have a power of 80% to detect a
JID 2000;182 (October)
Ascariasis and CVD 103-HgR Vaccine
difference in seroconversion rates between the 2 groups of 25%, with 61 subjects in each group [29]. This estimate was based on the findings of a previous study comparing seroconversion rates after vaccination with 5 3 10 8 cfu of CVD 103-HgR between subjects of low and high economic status in Peru [18]. Because the expected losses to follow-up were uncertain, the study size was increased to 120 subjects in each group. Randomization codes were kept in a sealed envelope at the Pharmaceutical Development Service, National Institutes of Health. The codes were broken only after all analyses had been completed and the data had been doubleentered into a computer. Vibriocidal antibody responses were expressed as geometric mean titers (GMTs) and rates of seroconversion (4-fold increase in titer). Helminth egg counts were loge transformed before analysis. Differences in frequencies were calculated with the x2 test or Fisher’s exact test, as appropriate. Comparison of means was performed with Student’s t test or the paired t test, as appropriate. After we adjusted for baseline vibriocidal antibody levels, differences in postvaccination vibriocidal antibody levels between treatment groups and between blood groups and interaction between treatment and blood groups were studied by analysis of covariance. The relationship between rates of seroconversion in the 2 intervention groups and blood group was evaluated by multiple logistic regression, after we had controlled for potential confounders.
Results Study population. Stool samples from a total of 1333 children from 27 schools in Canton Pedernales were collected from a population of 1650 children 6–13 years old, of whom 265 (19.8%) had infection intensities with A. lumbricoides 11000 epg in 2 consecutive samples. All these subjects met the other inclusion criteria. Because all these children came from small, widely dispersed, and poorly accessible rural communities, it was decided for logistic reasons to exclude communities with !5 eligible children; a total of 32 children were excluded in this way. A total of 233 children were recruited from 17 communities; 118 were assigned to the placebo group and 115 to the albendazole group. There were significant losses to follow-up during the study, and paired serum samples (>1 prevaccination sample and 10 days after vaccination) were obtained from a total of 139 subjects: 64 (54%) children in the placebo group and 75 (65%) in the albendazole group. Losses were due to adverse circumstances relating to heavy flooding because of the exceptionally heavy El Nin˜o rains that occurred; the rains caused the closing of all schools, made many study communities inaccessible by road, and resulted in temporary population migrations from the communities to farms in the hills. Thus, at several critical study times (e.g., the time of vaccination and day 10 after vaccination), the study team was either unable to gain access to communities or had to identify and visit by foot each household in these widely dispersed rural communities. Considerable efforts were made by the study team to establish and maintain good relations with the study communities, and losses to follow-up because of noncompliance were very few
1201
(!10%). The baseline characteristics of the study group are shown in table 1. There were no differences between the 2 groups with respect to age, sex, hematocrit levels, anthropometric measurements, and blood-group distribution. The mean pretreatment infection intensities with A. lumbricoides in the placebo and albendazole groups were 10,372 epg (95% confidence interval [CI], 7960–13,514) and 11,422 epg (95% CI, 9021–14,462), respectively. The prevalence of other intestinal helminths in the albendazole and placebo groups, respectively, were as follows: T. trichiura (62.7% vs. 79.7%; P p .03), hookworm (2.7% vs. 7.8%; P 1 .1), S. stercoralis (2.7% vs. 0%; P 1 .1), and Hymenolepis nana (4.0% vs. 0%; P 1 .1). Infection intensities of all these helminths were low. There were no differences at baseline with respect to age, anthropometric indexes, and helminth infection prevalences and intensities between those lost to follow-up and those included in the final analysis (data not shown). Response to anthelmintic treatment. Study recruits received either 2 doses of 400 mg of albendazole or placebo tablets on study days 0 and 30. The response to anthelmintic treatment for A. lumbricoides and T. trichiura is shown in figure 1. Because stool samples were not collected from all subjects at each time point after day 0, geohelminth infection intensities are shown as percentage changes compared with pretreatment levels. Immediately before vaccination (study day 37), the infection intensity or prevalence of A. lumbricoides infection did not alter significantly in the placebo group, although
